Drought in Africa Caused Delayed
Arrival of European Songbirds
A. P. Tøttrup,1* R. H. G. Klaassen,2,3 M. W. Kristensen,4 R. Strandberg,2 Y. Vardanis,2
Å. Lindström,2 C. Rahbek,1 T. Alerstam,2 K. Thorup4
dvancement in phenological events such
as spring arrival of songbirds at their breeding grounds is well documented (1). Despite this, the arrival in 2011 at northern European
breeding grounds of several trans-Saharan migratory bird species was among the latest documented
since 1950 (Fig. 1A). Local nonbreeding conditions can affect subsequent life-cycle stages at the
level of individual birds (2). Here, we confirm this
by showing a link of causality between the drought
at the Horn of Africa (3) and the migratory behavior
of individual birds resulting in subsequent delayed
arrival at breeding grounds thousands of km away.
Photoperiod (4) and local environmental conditions (5) control timing of life-history events in
migratory animals. Timing may also be affected by
events during migration, e.g., on stopover sites.
We followed individual movements of two longdistance migratory songbird species [18 red-backed
shrikes (RS) and eight thrush nightingales (TN)]
during three consecutive annual cycles (2009 to
2012) by miniature light-level geolocators (6) with
A
the purpose of assessing the impacts of environmental conditions at different life-history stages.
Breeding area arrival was documented by using
63 years of monitoring data and related to en
route conditions by using remote-sensing data
(31-year vegetation index series).
The exceptional 2011 delay in spring arrival
could be traced back to a significant prolongation of
stopover time during northward migration at the
Horn of Africa, which was at that time affected by
extreme drought (Fig. 1C). Mean stopover time in
this region was longer during the drought year 2011
(18 days, SE = 1.72, n = 5 for RS and 29 days, SE =
4, n = 3 for TN) compared with the mean stopover
time for the adjoining years, 2010 and 2012, combined (9 days, SE = 0.90, n = 13 for RS and 21 days,
SE = 5, n = 5 for TN). This carried over to a corresponding delay in spring arrival and breeding
start in Europe (Fig. 1A and fig. S1) even apparent
at the individual level (two individuals tracked during two consecutive years; supplementary materials). There was no significant difference in the
timing of migration between 2010/2012 and 2011
up to the spring stopover period in northeast Africa.
The extent of the catastrophic drought (3) at
the refueling area (6) is illustrated by vegetation
indices (Fig. 1C and fig. S2). The drought probably caused food shortage, which slowed down
refueling rate and increased time needed to reach
the necessary fuel loads. Other southeast African
migrants were also delayed, whereas species not
relying on the Horn of Africa for refueling were not
(table S1). Thus, this demonstrates cross-season interactions in individual migratory animals (2).
Advances in tracking technology currently revolutionize migration research. Our results elucidate how new technology helps establish a direct
link between local climate during migration and
arrival/breeding conditions. Still, improvements
in small-animal tracking are required to provide
better precision globally (7).
Migration delay may have cascading effects on
breeding success, mortality, and timing of later lifehistory stages. Start of breeding of RS was delayed
in 2011, but population sizes and reproductive success were reported to be close to average (supplementary materials). However, the implications of
our results are that even very local stopover conditions during short time periods have crucial importance for the entire migration system, emphasizing
the challenges in conserving migration routes and
systems (8).
References and Notes
1. C. Parmesan, Annu. Rev. Ecol. Evol. Syst. 37, 637 (2006).
2. P. P. Marra, K. A. Hobson, R. T. Holmes, Science 282,
1884 (1998).
3. C. Wolff et al., Science 333, 743 (2011).
4. E. Gwinner, W. Wiltschko, J. Comp. Physiol. 125, 267 (1978).
5. C. E. Studds, P. P. Marra, Proc. Biol. Sci. 278, 3437 (2011).
6. A. P. Tøttrup et al., Proc. Biol. Sci. 279, 1008 (2012).
7. M. Wikelski et al., J. Exp. Biol. 210, 181 (2007).
8. D. S. Wilcove, No Way Home (Island, Washington, DC, 2008).
Acknowledgments: Danish Nature Agency (J.nr.SN 302-009)
and ethical committees in Malmö/Lund (M112-09) approved
the studies. Swedish Research Council, Centre for Animal
Movement Research, Villum Foundation, Swedish Environmental
Protection Agency and Danish National Research Foundation
financed this study. Contribution no. 259 from Ottenby Bird
Observatory. Tracking data are available at www.movebank.org
(digital object identifier: 10.5441/001/1.8jg3h6n7).
Downloaded from www.sciencemag.org on October 12, 2014
BREVIA
Supplementary Materials
www.sciencemag.org/cgi/content/full/338/6112/1307/DC1
Materials and Methods
Supplementary Text
Figs. S1 and S2
Table S1
References (9–17)
17 July 2012; accepted 23 October 2012
10.1126/science.1227548
Fig. 1. (A) Trends in first-arrival day of RS and TN at Ottenby, Sweden, during 1950 to 2012 (solid dots
are 2011 arrivals). (B) Migration revealed by year-round tracking of individuals during 2009 to 2012.
(C) Horn of Africa staging positions of individual RS (blue; n = 9 in 2010, n = 5 in 2011, n = 4 in 2012)
and TN (red; n = 1 in 2010, n = 3 in 2011, n = 4 in 2012). Background maps show eMODIS normalized
difference vegetation index (NDVI) (15 to 30 April); color range (red to green) indicates increasing
presence of green vegetation. The difference in number of northeast African staging days between
2010/12 and 2011 was significant for both species combined (P = 0.006, n = 26) as well as RS alone (P <
0.002, n = 18), but not the small sample of TN (P = 0.46, n = 8; Kruskal-Wallis).
www.sciencemag.org
SCIENCE
VOL 338
1
Center for Macroecology, Evolution, and Climate, University of
Copenhagen, Universitetsparken 15, DK-2100 Copenhagen, Denmark. 2Department of Biology, Lund University, Ecology Building, SE22362 Lund, Sweden. 3Montagu’s Harrier Foundation and Animal
Ecology Group, Centre for Ecological and Evolutionary Studies, Post
Office Box 11103, Groningen University, NL-9700 CC Groningen,
Netherlands. 4Natural History Museum, Center for Macroecology,
Evolution and Climate, University of Copenhagen, Universitetsparken
15, DK-2100 Copenhagen, Denmark.
*To whom correspondence should be addressed. E-mail:
[email protected]
7 DECEMBER 2012
1307